Paper Title
ACCURATE HAND POSE ESTIMATION USING DEPTH CAMERA AND IMU

Abstract
This paper presents a novel hand pose estimation system that integrates depth camera and inertial measurement unit (IMU) data to achieve high accuracy in detecting and tracking hand movements. The system leverages the Intel RealSense D435 depth camera and the Witmotion JY61P IMU sensor for precise hand position and orientation estimation. Hand detection is performed using Google's MediaPipe model, which identifies reference points on the hand from RGB and depth images, while the IMU provides additional pitch and roll angle measurements. Experimental results demonstrate the system's accuracy, with average positional errors of 2.33 mm and yaw angle errors of 0.057 degrees from the depth camera, and roll and pitch angle errors of 0.45 degrees and 0.34 degrees respectively from the IMU. Keywords - Hand pose Estimation, Depth Camera, Inertial Measurement Unit.